ex for a com- 
x information 
iry shape that 
object's index 
Its at the 2nd 
personal com- 
that most of 
especially the 
This implies 
ck interactive 
dex 
ism quite dif- 
lowing advan- 
;the structure 
'e efficient; 
vith irregular 
ind the query 
For instance, 
1eters along a 
lant areas be- 
e (refer to fig- 
  
  
  
  
  
  
  
  
  
  
  
00 10000 
ects | objects 
40 9. 20 
50 4. 70 
14 0. 18 
38 | 595. 40 
52 1. 95 
44 7.29 
16 0.19 
n second) 
     
—— 20m 
——-200m 
region query 
  
(4) Can control query precision level. Different 
applications require different precisions. For ex- 
ample, in a military GIS system, accurate results 
are often needed ;In an interactive tour information 
system , user can give a query condition on a touch 
screen with his or her finger. In this case, the 
query precision level at the first or second level is 
enough, therefore further computation is omitted, 
so the performance is improved. 
6 Conclusions 
Ladex is a new spatial index mechanism, it sup- 
ports index of spatial objects in irregular shape, 
and can control query precision level. The struc- 
ture is simple. The fundamental algorithms are of 
high performance. We integrated Ladex with B"- 
tree to improve the efficiency of spatial retrieval. 
In our Map Database Management System MDB, 
Ladex has changed the query format significantly. 
Users do not need to master complex SQL 
database language, they can use mouse to select 
objects or draw a query range in arbitrary shape on 
the screen, and immediately get the information of 
the desired objects. If users operate through touch 
screen, they can get what they want with their fin- 
gers. This is really " What you want, what you 
get" without any ad hoc language. 
Future work could focus on: (1) In application, 
find an approach to determine the optimal number 
of the first level blocks; (2) Study the relationship 
between the number of partition levels and spatial 
search efficiency; (3)Find a better way to expres 
feature information; (4) Extend Ladex mechanism 
to 3 and above dimensional space. 
References 
Aref, W. G. , and Samet,H. ;1991. Optimization 
strategies for spatial query precessing. Proc. of 
935 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996 
Beckmann,N. , Kriegel, H. P. , Schneider, R. ‚and 
Seeger,B. ,1990. The R *-tree: an efficent and ro- 
bust access method for points and rectangles. ACM 
SIGMOD, pp. 322—331. 
Faloutsos ,C. , Sellis, T. ,Roussopoulos ,N. ,1987. 
The R*-tree: A dynamic Index for Multi-Dimen- 
sional Objects ,Proc. of the 13th VLDB conf. , pp. 
507 — 518. 
Guttman, A., 1984. R-trees: a dynamic index 
structure for spatial searching. Proc. ACM SIG- 
MOD,pp. 47—57. 
Hinrichs, K. , and Nievergelt,J. , 1983. The grid- 
file: a data structure to support proximity queries 
on spatial objects. Proc. of the WG 83 pp. 100— 
113. 
Ibrahim, K. , and Christos, F. , 1994. Hilbert R- 
tree: An Improved R-tree Using Fractals,Proc. of 
the 20th VLDB Conf. pp. 500—509. 
Lu,H., and Ooi, B. C. , 1993. Spatial Indexing: 
Past and future. IEEE D ta Engineering ,16(3) 16 
"2t. 
Medeiros ,C. B. , and.Pires,F. ,1994. Database for 
GIS ,In SIGM OD Record. 
Monehouse,S. ,1992. The ARC/INFO Geographic 
Information System, Computers and Geosciences: 
An international Journal,18(4). 
Ousterhout,J. K. , Hamachi, G. T. ,Mayo,R. N. , 
Scott, W. S. ,and Taylor, G. S. , 1984. Magic: A 
VLSI Layout System. In 21st Design Automation 
Conference,pages 152—159, Alburquerque, NM. 
White ,M. ,1981. N-trees: Large Ordered Indexes 
for Multi-Dimensional Space. Application Mathe- 
matics Research Staff, Statistical Research Divi- 
sion, U.S. Bureau of the Census.